KR20120092581A - Advanced laser wavelength control - Google Patents

Abstract

Laser light wavelength control is provided by periodically predicting the next position of the light control prism using a model of the prism's operating characteristics. The predicted value is updated when the measured values of laser power and wavelength are obtained. However, since this prediction is made without waiting for the measurement, the prediction can be made more frequently than the laser discharge repetition rate, and the prism can be repositioned at discrete points in time that can occur more frequently than the laser discharge event. . This also reduces the performance degradation that can be caused by the presence of one pulse after the laser measurement, and the resulting laser control signal is applied.

Description

ADVANCED LASER WAVELENGTH CONTROL}

The present invention belongs to the field of laser light control, and more particularly to the field of controlling the wavelength of a laser light source that can be used in semiconductor lithography processes.

Photolithography is widely used in the semiconductor industrial process. Current photolithography typically uses a laser light source to provide a very narrow band of light pulses that illuminate a mask that exposes a photo-sensitive material on a silicon wafer. However, advances in semiconductor device parameters increase the dependence on the performance characteristics of the laser light source used. Increasing precision and speed of operation are increasingly required.

Referring now to FIG. 1, shown is a block diagram of a laser system 100 that can be used in current photolithography processes. The light source of the laser system 100 is a master oscillator (MO) 120.

As noted, when the MO chamber 120 fires, the resulting light is line narrowing module (LNM) 110 where light shines on the grating in LNM 110 through a prism (actually several prisms). Enter). This serves as an optical wavelength selector in that changing the position of the prism in the LNM 110 changes the wavelength of the laser light. This altered wavelength of laser light passes through the MO chamber 120 to the output coupler (OC) 130 and back to a stepper-scanner device (not shown), for example, responsible for handling and exposing the semiconductor wafer. do.

Output coupler 130 also delivers laser light output from MO chamber 120 to line-center analysis module (LAM) 170. LAM 170 is a wavelength sampler that measures the wavelength of light output from MO chamber 120. The laser light output measurement is transmitted from the LAM 170 to the control computer 160.

The control computer 160 uses the light output measurements to determine what changes are needed to reposition the prism in the LNM 110 to achieve the desired laser light output wavelength during subsequent laser discharge events. The position of the prism in the LNM 110 is controlled by the voltage applied to the piezoelectric transducer (PZT) 140 connected to the prism in the LNM 110. Therefore, control computer 160 determines what voltage should be applied to PZT 140 to achieve a new desired prism position.

The control computer 160 outputs a digital signal to the PZT drive electronics 150 indicative of the desired voltage change to be made to the prism in the LNM 110. The PZT drive electronics include a digital-to-analog converter (DAC) that converts the digital signal of the control computer 160 into an analog voltage signal, and an analog lowpass that reduces high frequency electrical noise and amplifies the analog DAC voltage signal. Include a filter. This analog voltage signal then repositions the prism in the LNM 110 causing a change in the wavelength of the light output passing through the output coupler from the MO chamber 120 at a subsequent laser discharge event from the PZT drive electronics 150. Delivered to PZT 140.

This process continues when the stepper-scanner requests an additional sequence or burst of light pulses from the laser system at a particular pulse rate, start time, and wavelength.

As the photolithography process has evolved over the years, various challenges have arisen. For example, a reduction in semiconductor feature size has resulted in a reduction in the desired wavelength of the laser light source to maintain the desired focus. Reducing the wavelength requires much higher precision for the output laser light.

Another challenge is generated by pulse periods that vary from 167 to 600 microseconds. At a high pulse repetition rate, the LAM 170 obtains the measurement, the control computer 160 calculates a new voltage value for the prism in the LNM 110 and passes the value to the PZT drive electronics 150, PZT drive electronics 150 analog-converts and filters new voltage values to PZT 140, and PZT 140 changes the wavelength of light from MO chamber 120 before subsequent laser light pulses occur. Very little time is available to change the prism position.

Referring now to FIG. 2, shown is a timing diagram of the sequence. Two laser discharge events are shown from this sequence of pulses in one burst, the first discharge event occurring at time t ₀ and the second discharge event occurring at time t ₃ . After the first laser discharge event at time t ₀ , the resulting output wavelength is measured by LAM 170 (see FIG. 1 again). The delay of LAM 170 processing to measure the output wavelength and provide it to control computer 160 is shown in the figure with a LAM delay from time t ₀ to time t ₁ . The time at which the new control signal is then applied to the prism in the LNM 110 is shown at time t ₂ . The delay between the times t ₁ and t ₂ is calculated from the PZT driving electronics 150 to the PZT 140 to calculate the new voltage and reposition the prism until the laser discharge at t3. Time taken by computer 160.

The LAM 170 measurement delay is very fixed in that it depends on the LAM 170 processing time and the transfer time from the LAM 170 to the control computer 160. However, the laser discharge rate (time between t ₀ and t ₂ ) is determined by the stepper-scanner. As a result, due to the faster laser discharge rate, the LAM 170 measurement delay can potentially be long enough for subsequent laser discharge events to occur before a new control signal is applied. If it does, performance degradation may occur because the new control signal is based on measurements from a previous laser discharge event (ie, which may be one pulse later), rather than on the most recent laser discharge event. Therefore, there is a need for improved laser control that can operate at higher laser discharge rates.

Another problem with the known approach to laser control is the various disturbances in the laser system 110 that make it more difficult to precisely determine how to position the prism in the LNM 110. Therefore, there is a further need for improved laser control that can handle various disturbances occurring in laser systems.

Improved systems and methods for controlling the wavelength of a laser without relying on receiving laser light measurements prior to repositioning the prism are shown and described herein with reference to a number of specific embodiments.

As one embodiment, a laser wavelength control method is disclosed. The method includes predicting the position of the prism in the laser system at a first discrete interval, and receiving a measurement of the output wavelength of the master oscillator chamber in the laser system at a second regular interval longer than the first regular interval. For each first discrete interval, if the output wavelength measurement was not received during the first discrete interval, the control computer calculates the control voltage for the prism at the control computer using the predicted position of the prism, and the output wavelength measured value is determined as the first discrete interval. If received during an interval, the output wavelength measurement is used to update the predicted position of the prism, and the control voltage for the prism is calculated at the control computer using the updated predicted position of the prism, and the calculated control voltage from the control computer. Outputting the prism to an electronic device for positioning the prism; The.

In another embodiment, a laser wavelength control method is provided. The method includes predicting the position of the prism in the laser system at a first discrete interval, and receiving a measurement of the output wavelength of the master oscillator chamber in the laser system at a second discrete interval longer than the first regular interval. And if the output wavelength measurement was not received during the first discontinuous interval, the control computer computes the control voltage for the prism using the predicted position of the prism, the output wavelength measurement being the first discontinuous interval. If so, if the laser discharged again after receiving the output wavelength measurement, updating the previous prism position prediction using the output wavelength measurement, and creating a new prism position prediction based on the updated previous prism position prediction. , And new prism position estimates Calculating the control voltage for the prism at the control computer, and if the laser did not discharge again after receiving the output wavelength measurement, using the output wavelength measurement to update the predicted position of the prism and the updated prism position Calculating the control voltage for the prism at the control computer using the predicted value, and outputting the calculated control voltage from the control computer to the electronics for positioning the prism.

In another embodiment, a prism control method is provided. The method predicts the position of the prism at a first discrete interval using a prism motion model, and receives a measurement of the output wavelength of the master oscillator chamber controlled by the prism at a second discrete interval longer than the first regular interval. Calculating an control voltage for the prism using the predicted position of the prism if the output wavelength measurement was not received during the first discontinuous interval during each first discontinuous interval, the output wavelength measurement during the first discontinuous interval If received, updating the predicted position of the prism using the output wavelength measurement and calculating the control voltage for the prism using the updated predicted position of the prism, positioning the prism with the calculated control voltage. Outputting to an electronic device.

Various embodiments are disclosed in the detailed description below and in the accompanying drawings.
1 is a block diagram of an exemplary laser system that may be used in one embodiment.
2 is a timing diagram illustrating a prior art event sequence.
3 is a timing diagram illustrating an event sequence according to one example embodiment.
4 is a flowchart of one embodiment of the method.
5 is a timing diagram illustrating a determination of whether the laser will discharge again before a sequential voltage update event, according to one embodiment.
6 is a timing diagram illustrating prediction for a sequential voltage update event, according to one embodiment.
7 is a timing diagram illustrating prediction for the next laser discharge in accordance with one embodiment.
8 is a timing diagram illustrating updating a predicted value for a next voltage update event with a new laser light measurement, according to one embodiment.
9 is a timing diagram illustrating updating a predicted value for a previous laser discharge event and creating a new predicted value for a next voltage update event based on the updated predicted value for a previous laser discharge event, according to one embodiment.

Improved systems and methods are provided for controlling the wavelength of a laser that can be used in semiconductor photolithography. Rather than waiting for laser light measurements before determining how to reposition the prism in the laser system, this improved system method uses known models of disturbance behavior and prismatic behavior to periodically determine the next position of the prism in the laser system. Predict. The period of this prediction is typically faster than the period of the laser discharge event, so that the interval between predictions is typically shorter than the interval between laser discharge events. This removes the prior art dependency of having to receive laser light measurements before repositioning the prism. The measurement of the laser output wavelength is still obtained when a laser discharge event occurs and used to update or improve the prism position prediction. Whether updated or not, the predicted prism position is used to calculate the control signal used to drive the PZT controlling the prism position. In this way, prediction is made periodically, control signals are calculated periodically, and the prism is repositioned periodically accordingly.

The improved laser control systems and methods described herein can be thought of as "fast-discrete / low-speed discrete filters". Since the prediction of the prism position is made without waiting for the output wavelength measurement, the prediction can be made more frequently than the laser discharge repetition rate and the resulting output wavelength measurement (so-called "low speed-discrete"). As a result, the prism can be repositioned at discrete points in time that can occur more frequently than laser discharge events. Also, separating the prediction made when a laser discharge event occurs means that the frequency of updating the prism control is no longer limited by the time it takes to measure the laser output wavelength (“LAM delay” in FIG. 2). . It can also reduce performance degradation due to one pulse behind the laser measurements and the resulting laser control signal application.

Referring now to FIG. 3, there is shown a timing diagram of the sequence. Periodic laser discharge events are indicated as occurring at times t ₀ and t ₄ , and wavelength measurement data is indicated as being available at time t ₂ . Thus, with the interval between the prism control signals shown as T _controllers , new prism control signals applied as indicated at times t ₁ , t ₃ , t ₅ and t ₆ occur more frequently. This prism control signal is the result of new prism predictions made during each cycle of the control signal application. In addition, such prism prediction is made without waiting for the wavelength measurement data to become available. However, as mentioned above, when new wavelength measurement data is available, it is used to update the prism position prediction. As shown, the frequency at which the new control signal is applied occurs faster than the frequency of the laser discharge event. Details, modifications, and alternatives of this operation are further described elsewhere herein.

Referring again to FIG. 1, a block diagram of the laser system 100 will be used to describe one embodiment of the apparatus of the present invention.

The control computer 160 performs periodic prediction of the prism position within the LNM 110 using a model of the prism movement of the laser system 100, as described in more detail elsewhere herein. The control computer 160 uses the periodic prism position prediction to periodically output the digital control signal to the PZT drive electronics 150 representing the desired voltage to be applied to the prism in the LNM 110. The DAC of the PZT driving electronics 150 converts an analog voltage signal into a digital signal of the control computer 160, and the low pass filter of the PZT driving electronics 150 reduces electrical noise of a high frequency, and the analog DAC voltage signal Amplify. This analog voltage signal is transferred from the PZT driving electronics 150 to PZT 140, which is an application of a new control signal. This voltage update event repositions the prism in the LNM 110 and subsequently causes a change in the wavelength of the light output from the MO chamber 120 through the output coupler at the next laser discharge event.

  As described above, the light source in the laser system 100 is the MO chamber 120. When the MO chamber 120 discharges, the resulting light is in the LMN 110 where the light acts as an optical wavelength selector through a prism (actually a plurality of prisms, as described in more detail elsewhere herein). Enter the line narrowing module (LNM) 110 shining on the grating. The laser light of the changed wavelength passes back through the MO chamber 120 to an output coupler that transmits the laser light output from the MO chamber 120 to the line center analysis module (LAM) 170. The LAM 170 measures the wavelength of the laser light and transmits the measured value to the control computer 160. When the control computer 160 receives the laser light output measurement from the LAM 170, the control computer 160 updates the prism position to update the predicted prism position, as described in more detail elsewhere herein. Use the measured value.

This process is repeated periodically during the subsequent prism control signal application / voltage signal update cycle, as shown, for example, in FIG. In addition, when the new wavelength measurement occurring at time t ₂ in FIG. 3 is used, control computer 160 will update the predicted prism position using the newly received measured data.

As mentioned above, the control computer 160 performs periodic prediction of the prism position in the LNM 110 using an operating model of the prism in the laser system 100. The operation model of the prism is based on knowledge of various characteristics of the physical operation of the prism in the laser system 100, and considers the vibration action of the disturbance and prism in the laser system 100. In a preferred embodiment, the model used, as an integral form, is as follows.

(A)

Where x (t + T) is the predicted state t seconds before the initial state x (t), e ^AT is the standard state transformation matrix for the linear system, and B is the standard input matrix.

Equation (A) thus makes predictions for future events based on the prediction or status of past events. For example, referring again to FIG. 3, equation (A) may be used to perform prism position prediction for time t ₆ based on a prediction previously made for time T ₅ , which is expressed in equation (A). T = t ₆ -t ₅ . It should be understood that the initial condition is assumed for the initial prediction (e.g., the initial condition at laser boot is a predetermined constant, and the initial condition at the start of the sequence of laser pulses is the last state and setpoint change from the end of the previous laser pulse sequence and Equal to the sum of operations between any burst).

It should also be understood that the equation takes into account the disturbance of the wavelength within the laser system 100. For example, burst transients and steady-state disturbances cause unwanted laser light wavelength variations. Starting-of-burst transients are typically seen at measured wavelengths, which are a function of laser chamber acoustics and are caused by shock waves known as changes in laser discharge rate that complicate matters further. It is thought to be. This startup burst transient is modeled as a random walk process. Steady state disturbances appear as sinusoids in the wavelength signal and are so modeled. This sinusoidal steady state disturbance is thought to be caused by a fan, among other vibration sources within the laser system 100. Both these disturbances and any other disturbances identified within the laser system 100 may be considered in the model used for prism position prediction.

As mentioned above and described in more detail elsewhere herein, the prism position prediction value is updated when a new laser light measurement is received. The updating of the prism position prediction value is performed by the following equation.

(B)

Where

k indexes the laser discharge event.

I is an identity matrix of appropriate size (n rows x n columns, where n is the number of terms in the state vector x).

L _k is a gain matrix that captures the tradeoff between "trusting new data with trusting previous predictions".

C _k is the standard mapping from the predicted state to the predicted output. For example, y = C _x .

x (k | k) is the predicted value updated when the last data at index k is given.

-x (k | k-1) is the "old" prediction of the state at the laser discharge index k given the data from the previous laser discharge event, k-1.

Therefore, Equation (B) updates the previously made prediction as described in more detail elsewhere herein.

Referring now to FIG. 4, a flowchart of one embodiment of the present invention is shown. In the figures, it can be seen that the method is logically divided into three sections, but it is to be understood that the described functionality may be implemented in alternative configurations. The measurement update section 401 relates to an update operation that occurs when a new laser light measurement is received. Calculation control section 402 relates to the operations that occur when computing a new control signal using the prism position prediction. The time update section 403 is primarily concerned with operations that occur to make predictions for subsequent voltage update events. Each of these will now be described in more detail in the context of one embodiment.

For example, the present process running on the control computer 160 of FIG. 1 begins at the measurement update section 401 when the present process is started periodically, for example by a timer in the control computer 160.

If a new wavelength measurement is not available at decision block 405, the process leaves the measurement update section 401 and enters the calculation control section 402. In the calculation control section 402, a previously made new prism position prediction value is used in step 430 to calculate the value for the coarse wavelength prism in the LNM 110. As noted, the typical configuration of LNM 110 includes one or more prisms, one of which is a coarse wavelength prism used to maintain the PZT near its midrange voltage, while the other is used for fine wavelength control. In the present specification, a prism commonly referred to as a prism. Although not shown in FIG. 4, new values for the coarse wavelength control prism are output from the control computer 160 to properly drive the electronics.

In step 435, the previously generated next prism position prediction is used to calculate the delta PZT voltage. The delta or change in PZT voltage is calculated according to the equation below.

(C)

m indexes the control event (ie voltage update event).

τ c is the controller sample period, thus creating an x (mτ c) state at the (arriving) voltage update time.

k is a state feedback matrix calculated to minimize the weighted sum of control energy and wavelength performance.

This delta PZT voltage is then used in step 440 to calculate a new PZT control signal. The new PZT control signal is calculated according to the equation below.

(D)

V ((m−l) τc) is the PZT control signal from the previous update.

Δu (mτc) is the result of equation (C).

V (mτc) is the new PZT control signal.

If the context is clear, it should be understood that a signal such as V (mτc) will be denoted V _m for brevity.

Although not shown in FIG. 4, the PZT control signal is output to, for example, the PZT drive electronics 150 of FIG. 1 for application to a prism.

The process exits the computational control section 402 and, as described with reference to FIG. 5, the time determined at step 445 whether the laser will discharge again before a subsequent voltage update event (the subsequent application of a new control signal to the prism). Continue by entering update section 403.

As can be seen in FIG. 5, new control signals, denoted by V _{m -i} V _m and V _{m + i} , will be applied periodically. However, the current process is marked as starting at time C _m , and there are two scenarios to consider whether the laser will discharge before a subsequent voltage update event.

Before describing these two scenarios, the process this time point at, i.e., from C _m, the subsequent voltage update event is not the application of the next voltage update event of the new control signal V _m a new control signal V _{m + i} of the application Should be understood. This difference between the next voltage update event and the subsequent voltage update event is simplified as a function in which the periodic process is within that cycle. Likewise, the subsequent voltage update event becomes the next voltage update event of the next cycle.

As already mentioned, there are two scenarios where the laser discharges before a subsequent voltage update event at the point at time C _m . The first scenario is when the laser system has already received a trigger signal from the stepper-scanner system, as depicted at the point in time T _{k + i} in the figure, for the MO chamber 120 to discharge again. This received trigger signal is then used to determine the time of the resulting laser discharge event, shown as LF _{k + i} in the figure, to determine if the next laser discharge event will occur (in this example) before the subsequent voltage update event V _{m + i} . It can be used in combination with a known laser trigger delay (ie, a known lag between the scanner trigger signal and subsequent light generation) to estimate.

The second scenario is when the laser has not yet received a trigger signal from the stepper-scanner system, which is depicted by a trigger received at a later point in time T _{k +1} in the figure, which is after the current time C _m . If you have not yet received the trigger signal, the present process is that instead of a periodic trigger signal and a known trigger delay, and V _{m + i} to a not based on the trigger delay, the drawing shown as LF _{'k + i,} which receives one or more previous , Estimate the laser discharge event. The process then determines whether this estimated laser discharge event LF ' _{k + i} will occur (in this example) before the subsequent voltage update event V _{m + i} .

The process now determines whether, under two possible scenarios, the laser will discharge before the next voltage update event V _{m + i} .

Referring again to FIG. 4, if the result of decision block 445 is that the laser does not discharge prior to a subsequent voltage update event, then at step 455, a prism position prediction for a subsequent voltage update event is made.

Creating a prism position prediction value for a subsequent voltage update event, performed at step 455, is now described with reference to FIG. 6. As before, the process is at the current time C _m . As indicated by the arrows in the figure, a prediction is made for a subsequent voltage update event designated V _{m + i} , which prediction is based on a previously made prediction for the next voltage update event designated V _m . In a preferred embodiment, prediction for subsequent voltage update events is made using equation (A) described elsewhere herein.

Referring again to FIG. 4, the process then ends the operation of one cycle in the time update section 403.

Predictions are now made for subsequent voltage update events, and the sequence of decision block 405 determines that there are no new measurements made available by calculating new control signals at steps 430, 435, and 440, and determines It should be understood that determining that block 445 will not discharge the laser before another subsequent voltage update event causes periodic prediction and repositioning of the prism at discrete points in time, as depicted in the timing diagram of FIG. 3. do.

Returning to decision block 445 and if it is determined instead that the laser will discharge before a subsequent voltage update event, in step 450, a prism position prediction for the next laser discharge event is made.

Predicting the prism position for the next laser discharge event, performed at 450, is now described with reference to FIG. 7. As described above, the process is at the current time C _m . Further, in this example, and as described with reference to FIG. 5, it can now be seen that the laser will discharge at LF _{k + i} after the next voltage control update V _m and before the subsequent voltage control update V _{m + i} . The process performs prism position prediction for the laser discharge event at LF _{k + i} based on the previously made prediction value for voltage update event V _m , as indicated by the prediction arrow in the current figure. In a preferred embodiment, this prediction for the laser discharge event at LF _{k + i} is made using equation (A) described elsewhere herein.

Referring back to FIG. 4, the process proceeds to step 455 where a prism position prediction is made for subsequent voltage update events, as described above. And, as described above, the process ends one cycle of operations in the time update section 403.

Referring again to the decision block 405, and to the beginning of the process of periodically starting in the measurement update section 401, if a new laser power measurement is received and available, the process uses the new measurement, Continue within measurement update section 401 by performing one or more updates to the predicted and / or new predicted values, which is further described below.

In decision block 410, if it is determined that the laser has not discharged again since the last time a new output measurement was received, the process continues to step 420 where the prediction for the next voltage update event is updated with the new measurement. .

Updating the predicted value for the next voltage update event with the new measured value, performed at step 420, is now described with reference to FIG. 8. As before, the process is at the current time C _m and the laser power measurement for the current laser discharge event LF _k is made available before the prediction for the subsequent voltage control event V _{m + i} (not shown) is made. do. This process uses the time remaining before the next control signal V _m is applied to update a previously made prediction for that voltage update event using the newly available laser power measurement. In this way, the newly available laser power measurements from the laser discharge event LF _k are used to update the previously made predictions for the new control signal application V _m . In a preferred embodiment, updating the predicted value for the next voltage update event is accomplished using equation (B) described elsewhere herein.

Referring back to FIG. 4, if the pre-made prism prediction value for the next voltage update event is now updated, the process leaves the measured value update section 401, as described above, with a coarse wavelength prism signal 430 or the like. It continues by entering the starting calculation control section 402. However, the current calculation steps 430, 435, and 440 use the updated prediction value for the next voltage update event, thus gaining gains from the newly received laser output wavelength measurement data.

Alternatively, referring back to decision block 410 of measurement update section 401, if it is determined that the laser has discharged again since the last time a new laser power measurement was received, the process may predict the previous laser discharge event. 415 is updated with this new measurement, and step 425, wherein the new prediction for the next voltage update event is based on the updated prediction for the previous laser discharge event, which will be described further below. .

As performed in step 415, updating the predicted value for the previous laser discharge event with the new measured value is now described with reference to FIG. As before, the process is at the current time C _m and the laser power measurements from the laser discharge event LF _k are now available. The process uses the new laser power measurement from the laser discharge event LF _k (laser discharge event LF _{k + i} (performed in step 450, in the previous process cycle, as described elsewhere herein)). Update the pre-made prediction value for. In a preferred embodiment, updating the predicted value for a previous laser discharge event is accomplished using equation (B) described elsewhere herein.

As performed in step 420, making a new prediction value for the next voltage update event based on the updated prediction value for the previous laser discharge event is now described with reference to FIG. As described above, the process is at the current time C _m . The updated prediction value for the previous laser discharge event LF _{k +1} from step 415 is used to predict for the next voltage update event shown in the figure that the new control signal Vm is applied, as indicated by the prediction arrow. do. In a preferred embodiment, making a new prediction value for the next voltage update event based on the updated prediction value for the previous laser discharge event is made using equation (A) described elsewhere herein.

Referring again to FIG. 4, if a new prediction is now made for the next voltage update event, the process leaves the measured value update section 401, as described elsewhere herein, to the coarse wavelength prism signal 430. And so on by entering computation control section 402, beginning with FIG. However, the calculation steps 430, 435, and 440 now use the newly generated predictions for the next voltage update event, thus benefiting from the newly received output wavelength measurement data and the known timing of the new laser discharge event.

Finally, the process cycle is completed by performing appropriate operations within the time update section 403 to predict for subsequent voltage update events, as described elsewhere herein.

Various different sequences of processing steps are described to reflect various scenarios that may occur to aid understanding.

As a first possible sequence, a control signal is calculated and output, and a prediction value for subsequent voltage events is generated. This scenario occurs when there is no new laser output wavelength measurement and the laser does not discharge before the next voltage update event. This sequence includes processing steps 430, 435, 440, and 455.

As a second possible sequence, a control signal is calculated and output, a prediction is made for the next laser discharge event, and a prediction value is generated for the subsequent voltage event. This scenario occurs when there is no new laser output wavelength measurement and the laser discharges before the next voltage update event. This sequence includes processing steps 430, 435, 440, 450 and 455.

As a third possible sequence, the predicted value for the next voltage update event is updated, the control signal is calculated based on the updated next voltage update event, and the prediction for the next voltage update event is made. This scenario occurs when the new laser output wavelength measurement becomes available, the laser did not discharge again after the last reception of the new measurement and the laser did not discharge before the next voltage update event. This sequence includes processing steps 420, 430, 435, 440, and 455.

As a fourth possible sequence, the predicted value for the next voltage update event is updated, the control signal is calculated based on the updated next voltage update event, a prediction is made for the next laser discharge event, and for a subsequent voltage update event. Predictions are made. This scenario occurs when the new laser output wavelength measurement becomes available, the laser did not discharge again after the last reception of the new measurement and the laser discharged before the next voltage update event. This sequence includes processing steps 420, 430, 435, 440, 450 and 455.

As a fifth possible sequence, the prediction value for the previous laser discharge event is updated, and the prediction for the next voltage update event is based on the updated laser discharge event, and the control signal is based on the prediction value for the next voltage update event. And a prediction is made for subsequent voltage update events. This scenario occurs when a new laser output wavelength measurement becomes available, the laser discharges again after the last reception of the new measurement and the laser does not discharge before the next voltage update event. This sequence includes processing steps 415, 425, 430, 435, 440, and 455.

As a sixth possible sequence, the prediction value for the previous laser discharge event is updated, the prediction for the next voltage update event is based on the updated laser discharge event, and the control signal is based on the prediction value for the next voltage update event. Is calculated, a prediction is made for the next laser discharge event, and a prediction is made for a subsequent voltage update event. This scenario occurs when a new laser output wavelength measurement becomes available, the laser has not discharged again since the last reception of the new measurement, and the laser will discharge before the next voltage update event. This sequence includes processing steps 415, 425, 430, 435, 440, 450 and 455.

The embodiments described herein are for the purpose of illustrating the invention. These embodiments of the present invention are described with reference to the drawings, and various modifications and variations of the described method and / or specific structure may be made by those skilled in the art. All such modifications, adjustments, or variations, and techniques advanced through such teachings, in accordance with the teachings of the invention are considered to be within the spirit and scope of the invention. Therefore, it is to be understood that the description and drawings are not to be interpreted in a limiting sense, and that the invention is not limited to the embodiments described in any way.

As one possible alternative, it should be understood that the improved laser control methods and apparatus described herein may be used for bandwidth control of laser light in a manner similar to that described herein for wavelength control of laser light.

It is also to be understood that alternative sequences and equations may be used within the spirit and scope of the invention described herein.

Similarly, control computer 160 may include a processor and memory, including a personal computer, server, or other processing system, which resides on a computer readable medium or executes software instructions for performing the above-described instructions. It can be any computing system provided. Alternatively, control computer 160 may be any dedicated hardware, such as an application specific semiconductor (ASIC) or other hardwired device, with or without firmware, specifically configured to perform the operations described above.

Claims (15)

As a laser wavelength control method,
Predicting the position of the prism in the laser system at a first discrete interval;
Receiving a measurement of an output wavelength of a master oscillator chamber in the laser system at a second discrete interval,
During each first discrete interval:
If the output wavelength measurement has not been received during the first discontinuous interval, calculating a control voltage for the prism at the control computer using the predicted position of the prism;
If the output wavelength measurement was received during the first discontinuous interval, use the output wavelength measurement to update the predicted position of the prism and to control the prism using the updated predicted position of the prism. Calculating a voltage in the control computer, and
Outputting the calculated control voltage from the control computer to an electronic device for positioning the prism,
And wherein the second regular interval is longer than the first regular interval. The method of claim 1, wherein predicting the position of the prism at the first discontinuous interval uses an operation model of the prism. 3. The method of claim 2, wherein the motion model of the prism is based on physical motion of the prism and disturbances in the laser system. The method of claim 2, wherein the model is in the form of an integral

Laser wavelength control method characterized by the above-mentioned. The method of claim 1, wherein updating the predicted position of the prism using the output wavelength measurement is as follows:

Laser wavelength control method characterized in that for using. As a laser wavelength control method,
Predicting the position of the prism in the laser system at a first discrete interval;
Receiving a measurement of an output wavelength of a master oscillator chamber in the laser system at a second discrete interval,
During each first discrete interval:
If the output wavelength measurement has not been received during the first discrete interval, calculating a control voltage for the prism at a control computer using a predicted prism position;
If the output wavelength measurement was received during the first discontinuous interval,
When the laser is discharged again after receiving the output wavelength measurement, updating a previous prism position prediction value using the output wavelength measurement value, creating a new prism position prediction value based on the updated previous prism position prediction value. Calculating at the control computer the control voltage for the prism using the new prism position prediction value, and
When the laser does not discharge again after receiving the output wavelength measurement, updating the predicted position of the prism using the output wavelength measurement, and using the updated prism position prediction value Calculating the control voltage with respect to the control computer, and
Outputting the calculated control voltage from the control computer to an electronic device for positioning the prism,
And wherein the second regular interval is longer than the first regular interval. 7. The method of claim 6, wherein the first discrete interval is shorter than the second discrete interval. 7. The method of claim 6, wherein the step of making the predicted value of the position of the prism a first discontinuous interval uses an operating model of the prism. 10. The method of claim 8, wherein an operating model of the prism is based on the physical operation of the prism and the disturbance in the laser system. As a prism control method,
Predicting the position of the prism at a first discrete interval using a prism motion model;
Receiving a measurement of the output wavelength of a master oscillator chamber controlled by the prism at a second discontinuous interval,
During each first discrete interval:
If the output wavelength measurement has not been received during the first discontinuous interval, calculating a control voltage for the prism using the predicted position of the prism;
If the output wavelength measurement was received during the first discontinuous interval, updating the predicted position of the prism using the output wavelength measurement and using the updated predicted position of the prism to the prism. Calculating a control voltage for, and
Outputting the calculated control voltage to an electronic device for positioning the prism;
And wherein the first regular interval is longer than the first regular interval. 11. The method of claim 10, wherein the first discontinuous interval is shorter than the second discontinuous interval. 11. The method of claim 10, wherein the step of making the predicted value of the position of the prism a first discontinuous interval uses a motion model of the prism. 13. The method of claim 12, wherein the motion model of the prism is based on a physical motion of the prism and a disturbance in the laser system. 13. The system of claim 12, wherein the model is in its integral form.

Prism control method characterized in that. 11. The method of claim 10, wherein updating the predicted position of the prism using the output wavelength measurement is:

Prism control method characterized in that using.

Images